Profiling Student Behavior in a Blended Course

Closing the Gap between Blended Teaching and Blended Learning

Nynke Bos

and Saskia Brand-Gruwel

University of Amsterdam, Faculty of Social and Behavioural Sciences,

P.O. Box 19268, 1000 GG Amsterdam, The Netherlands

Open University of the Netherlands, Faculty Psychology and Educational Sciences,

P.O. Box 2960, 6401 DL Heerlen, The Netherlands

Keywords: Blended Learning, Blended Teaching, Learning Analytics, Recorded Lectures, Formative Assessment,

Individual Differences, Cluster Analysis, Learning Dispositions.

Abstract: Blended learning is often associated with student-oriented learning in which students have varying degrees

of control over their learning process. However, the current notion of blended learning is often a teacher-

oriented approach in which the teacher identifies the used learning technologies and thereby offers students

a blended teaching course instead of a blended learning course (George-Walker & Keeffe, 2010). A more

student-oriented approach is needed within educational design of blended learning courses since previous

research shows that students show a large variation in the way they use the different digital learning

resources to support their learning. There is little insight into why students show distinct patterns in their use

of these learning resources and what the consequences of these (un)conscious differences are in relation to

student performance. The current study explores different usage patterns of learning resources by students in

a blended course. It tries to establish causes for these differences by using dispositional data and determines

the effect of different usage patterns on student performance.

1 INTRODUCTION

When discussing learning technologies, there seems

to be consensus about its positive impact on

education. Phrases as ‘new potential’, ‘rapid and

dramatic change’ and ‘fast expansion’ are frequently

used when describing new learning technologies.

This is no different for blended learning as the

abovementioned phrases are used to characterize

current developments within the blended learning

domain (Henderson et al., 2015).

The definition of blended learning is not clearly

defined and can relate to combinations of

instructional methods (e.g. discussions, (web)

lectures, simulations, serious games or small

workgroups), different pedagogical approaches (e.g.

cognitivism, connectivism), various educational

transfer methods (online and offline) or it can relate

to various technologies used (e.g. e-learning,

podcasts or short video lectures (Bliuc et al., 2007;

Porter et al., 2016).

The common distinction lies in the two different

methods used within the learning environment: face-

to-face (offline) versus online learning activities.

Blended learning is often associated with

student-oriented learning, in which students have

varying degrees of control over their own learning

process. Blended learning could contribute to the

autonomy of the students in which they have more

control over their learning path and this autonomy

should encourage students to take responsibility for

their own learning process (Lust et al., 2013). This

approach towards blended learning is in line with a

constructivist pedagogical model and is believed to

assist in a flexible learning environment where

student autonomy and reflexivity is strengthened

(Orton-Johnson, 2009). However, in most cases the

design of blended learning is mostly aimed at

putting technology into the learning environment

without taking into account how that technology

contributes to the learning outcomes (Verkroost et

al., 2008) or encourages student autonomy and

reflexivity. The current notion of blended learning is

often a teacher-oriented approach in which the

teacher determines the learning technologies without

considering how these learning technologies

Bos, N. and Brand-Gruwel, S.

Proﬁling Student Behavior in a Blended Course - Closing the Gap between Blended Teaching and Blended Learning.

In Proceedings of the 8th International Conference on Computer Supported Education (CSEDU 2016) - Volume 2, pages 65-72

ISBN: 978-989-758-179-3

contribute to flexible learning, student autonomy and

course performance. This so called ‘blended

teaching’ approach (George-Walker and Keeffe,

2010) lacks a focus on students. To improve the

educational design of blended learning, a focus on

students is needed so students can choose the ‘right’

learning technologies to be suit their own learning

path. What the teacher determines as the ‘right’

technology does not necessary match with the

perspective of the learner and will not automatically

lead to more student-oriented learning (Oliver and

Trigwell, 2005) or encourages student autonomy and

reflexivity.

When blended learning design focuses on

students and their choices to use the ‘right’ learning

technologies, there is a large variety in the choices

students display when using the different learning

resources to support their learning. Students either

heavily rely on a single preferred supporting

technology (Inglis et al., 2011) do not use the

technology at all (Lust et al., 2011) or apply it in

such a way to substitute for the face-to-face

activities (Bos et al., 2015), thereby de facto creating

their own online course. One blended teaching

course can thereby lead to different blended learning

courses. There is little insight into why students do

or do not use certain learning technologies and what

the consequences of these (un)conscious choices are

in relation to student performance, although research

suggests that goal-orientation (Lust et al., 2013),

approaches to learning (Ellis et al., 2008) may be an

important predictor of frequency and engagement of

use.

Several studies conducted a cluster analysis

based on the use of these different learning resources

to identify different usage patterns. For example

Lust et al., (2013) found four different clusters that

reflect differences in the use of the digital learning

resources: the no-users, the intensive-active users,

selective users and intensive superficial users.

Similarly another study (Kovanović et al., 2015)

found, also based on cluster analysis, several

different user profiles based on the use of digital

learning resources and suggest that these differences

might be related to differences in students’

metacognition and motivation.

One of the advantages of blended learning is that

the learning activities take place in an online

environment, which easily generates data about

these online activities. The methods and tools that

aim to collect, analyse and report learner-related

educational data, for the purpose of informing

evidence-based educational design decisions is

referred to as learning analytics (Long and Siemens,

2011). Learning analytics measures variables such as

total time online, number of online sessions or hits

in the learning management systems (LMS) as a

reflection of student effort, student engagement and

participation (Zacharis, 2015). Learning data

analysis from students in a blended learning setting

provides the opportunity to monitor students’ use of

different learning technologies throughout the course

and might provide insight in the gap between the

education design of the course and the different

learning paths of students.

To better understand student behaviour in a

blended learning setting, learning data analysis

needs to be complemented with a set of indicators

that goes beyond clicks and durations of use. One

solution is to combine data from online learning

activities with learning dispositions, values and

attitudes, which should be measured through self-

report surveys (Shum and Crick, 2012) such as the

Motivated Strategies for Learning Questionnaire

(MSLQ) (Pintrich et al. 1991). Learners’

orientations towards learning—their learning

dispositions—influence the nature of how students

engage with new learning opportunities. Someone

who is able to self-regulate his or her own learning

process is more likely to use a deep approach

towards learning (Vermunt, 1992) Students who use

an external regulation strategy are more likely to use

a surface approach towards learning. So, adding

learning dispositions to data collected from online

learning activities could provide better

understanding of students’ regulation strategies and

their use of learning technologies, and subsequently

explain differences in student performance. Indeed,

preliminary research shows that dispositional data

adds to the predictive power of learning analytics

based on prediction models (Tempelaar et al., 2015).

To close the gap between blended teaching and

blended learning a deeper understanding of the

causes of individual differences of the use of

learning resources is needed so the educational

design process can be optimized. Dispositional data

could be used to determine if differences in students

their metacognition and motivation can explain

differences in the use of learning resources and what

the consequences of these differences are for student

performance.

This research aims to answer the following

questions:

Q1: Which differences in the use of learning

resources can we distinguish?

Q2: Can these differences be explained by

dispositional data?

Q3: Do these differences in the use of learning

CSEDU 2016 - 8th International Conference on Computer Supported Education

resources have an impact on student performance?

2 METHODS & MATERIALS

2.1 Participants

The participants were 516 freshmen law students

(218 male, 298 female, Mage = 22.1, SD = 4.9)

enrolled in a mandatory course on Contract Law.

Students repeating the course or taking the course as

an elective were removed from the results.

2.2 The Blended Learning Course

The course on Contract Law (CL) was an eight-

week course. The course had a regular outline for

each week.

On the first day of the week students were

offered a regular face-to-face lecture in which

theoretical concepts were addressed. These lectures

were university style lectures, with the instructor

lecturing in front of the class. The lectures were

recorded and made available directly after the

lecture had taken place and were accessible until the

exam had finished. If parts of the lectures were

unclear, students could use the recorded lectures to

revise these parts or revise the entire lecture if

needed.

The course consisted of 7 face-to-face lectures,

with a 120-minute duration and a 15-minute break in

half time. Lecture attendance was not mandatory.

During the week several small workgroups were

organized with mandatory attendance. Before these

workgroups, students had to complete several

assignments in the digital exercise book, which

contains additional study materials, supplemented

with three short essay questions. The students were

expected to have studied the digital exercise book

before entering the small workgroups. Responding

to the short essay questions was not mandatory, but

highly recommended by the instructor. In total there

were seven exercises that contained short essay

questions.

In the final segment of the week students were

offered a case-based lecture in which theoretical

concepts were explained with cases and specific

situations of Contract Law. These seven case-based

lectures were also recorded and made available

directly after the lecture had taken place and were

accessible until the exam had finished. All the

recorded lectures were made available through the

learning management system (LMS) Blackboard.

To finalize the week students could take a short

formative assessment in which the concepts of the

week were assessed. These formative assessments

contained multiple-choice questions in which

knowledge and comprehension were assessed.

Completion of these formative assessments was not

mandatory. In total there were seven formative

assessments available to students.

2.3 Measurement Instruments

The data collected from all the online activities

(recorded lectures, short essay questions, formative

assessments) was supplemented with the collection

of learning disposition data and attendance to the

face-to-face lectures: the regular lectures and the

case-based lectures.

2.3.1 Attendance to the Face-to-Face

Lectures

During the entire time frame of the lectures, student

attendance was registered on an individual level by

scanning student cards upon entry of the lecture hall.

The scanning continued until 15 minutes after the

lecture had started. The presence of the students was

registered for all fourteen lectures of the course,

seven regular lectures and seven case-based lectures.

Attendance to the regular lectures and the case-based

lectures was separately registered in the database.

2.3.2 Use of the Recorded Lectures

The viewing of the recordings was monitored on an

individual level and could be traced back to date,

time, amount and part of the lecture viewed. For

each lecture a separate recording was made, which

made it possible to track the viewing trends for that

specific recorded lecture.

2.3.3 Short Essay Questions

Since the digital exercise book was offered to

students through the LMS, answers given to the

short essay questions were also stored in the LMS.

These answers were not scored, students were

provided with model answers at the end of the week.

The LMS registered if a student had answered the

questions for that specific week.

2.3.4 Formative Assessments

For each formative assessment a log file within the

LMS was created to determine if a student

completed the formative assessment. For each

separate assessment a log file was created. The

Proﬁling Student Behavior in a Blended Course - Closing the Gap between Blended Teaching and Blended Learning

participation for the multiple choice and short essay

questions was stored separately.

2.3.5 Motivated Strategies for Learning

Questionnaire

The Motivated Strategies for Learning Questionnaire

(MSLQ) is a self-report instrument for students that

assess both student motivations and their

metacognitive ability to regulate learning (Pintrich et

al., 1991). The MSLQ contains 81 questions of

which 31 items determine a student’s motivational

orientation towards a course and 50 items to assess

metacognition. The motivational orientation can be

divided into six subscales: intrinsic goal orientation,

extrinsic goal orientation, task value, self-efficacy,

control beliefs and test anxiety. Metacognition can

be scored on nine subscales: rehearsal, elaboration,

organization, critical thinking, metacognitive self-

regulation, time and study environment, effort

regulation, peer learning and help seeking. For a

complete description of the MSLQ and each of its

subscales we refer to the manual of the MSQL

(Pintrich et al., 1991). For the purpose of this

research we used four motivation scales (intrinsic

goal orientation, extrinsic goal orientation, task

value and self-efficacy) and three metacognition

scales (critical thinking, metacognitive self-

regulation and peer learning) since these different

subscales can be, directly or indirectly, influenced

by their educational design within a blended learning

course.

The MSLQ was offered to students during the

first week of the course. In the second week a

reminder was sent out participants.

2.3.6 Final Grade

At the end of the course students took a summative

assessment, which consisted of 25 multiple-choice

questions and four short essay questions. Final

grades were scored on a scale from 1 to 10 with 10

the highest and 5.5 as a pass mark.

2.4 Data Analysis

To establish differences in the use of learning

resources a two-step cluster analysis with attendance

data, use of the recorded lectures, essay questions

and formative assessments was conducted. A two-

step cluster analysis determines the natural and

meaningful clusters that appear within an

educational blended setting. The two-step method is

preferred over other forms of cluster analysis when

both continuous and categorical variables are used

(Chiu et al., 2001) and when the amount of clusters

is not pre-determined.

Next a MANOVA between the different clusters

was conducted to determine significant differences

in student motivations and their metacognition

between those clusters (MSLQ). The MANOVA

was used to determine if dispositional data could

explain the existence of different clusters and

subsequently the differences in the use of learning

resources.

The last step in the data analysis was to conduct

an ANOVA with cluster membership as a factor and

with the final assessment as the dependent variable,

to determine if differences in the use of the learning

resources lead to significant differences in student

performance.

3 RESULTS

To determine the natural occurring patterns based on

the use of learning recourses a cluster analysis was

conducted. As can be seen in Table 1 the auto-

clustering algorithm indicated that four clusters was

the best model, because it minimized the Bayesian

Information Criterion (BIC) value and the change in

them between adjacent numbers of clusters.

Table 1: BIC changes in de auto-clustering procedure.

Number

Clusters

Schwarz's

Bayesian

Criterion (BIC)

BIC

Change

Ratio of

BIC

Changes

1 2217.94

2 1908.87 -309.06 1.33

3 1694.06 -214.82 1.54

4 1581.20 -112.86 2.27

5 1573.28 -7.92 1.11

6 1573.55 .28 1.04

a. The changes are from the previous number of clusters in the

table.

b. The ratios of changes are relative to the change for the two-

cluster solution.

Table 2 provides insight into the four different

clusters and their use of the learning recourses. For

each cluster the means of the use are presented as

well as the means for the entire population.

Students in cluster 1 hardly attend any of the

regular and case-based lectures; they hardly use the

short essay questions or the formative assessment

but show an average use of recordings of the

lectures. They seem to have a slight preference to

watch the recordings of the face-to-face lectures

over the case-based lectures. Students in cluster 2

CSEDU 2016 - 8th International Conference on Computer Supported Education

Table 2: Means of the learning data of the clusters.

  Lectures Case‐based

Lectures

Short

Essay

Formative

Assessments

Recorded

Lectures

(minutes)

Recorded

Lectures:Case

Based(minutes)

Cluster N M SD M SD M SD M SD M SD M SD

1 103 .28 .63 .12 .35 .53 .99 .49 .87 612 528 384 417

2 143 1.52 1.79 .17 .43 4.58 2.00 3.23 2.03 245 234 209 215

3 186 .39 .86 .21 .57 5.80 1.36 4.83 1.91 1013 421 740 493

4 84 4.80 1.87 2.43 1.83 5.68 2.07 4.93 1.93 301 363 371 326

Total 516 1.40 2.06 .54 1.19 4.39 2.57 3.53 2.45 604 516 462 447

attend some regular lectures, but they hardly attend

any of the case-based lectures. They show an

average activity on the use of the short essay

questions and formative assessments. Students in

cluster 3 hardly attend any of either type of face-to-

face lectures, but they compensate their lack of

attendance by watching the online recordings of both

types of lectures. They show an above average

activity on the assessments with a slight preference

for essay over multiple-choice questions. Students in

cluster 4 attend a well above average amount of the

face-to-face lectures. They also show an above

average activity on the assessments, but with a slight

preference for multiple-choice questions over short

essay questions. They show a modest use of the

recorded lectures.

To determine if the occurrence of these different

clusters could be explained by dispositional data, we

determined if there were significant differences

between the scores on the subscales of the MSLQ

between the four clusters. In total 103 students filled

out the MSLQ, which is 20% of the population. First

the reliability of the subscales of the MSLQ was

calculated. These results can be found in Table 3

The reliability of the subscales intrinsic goal

orientation, extrinsic goal orientation and

metacognitive self-regulation seems to be

insufficient. Therefore these subscales were

eliminated for further analysis. The low reliability of

the subscales is party caused by the limited items

that are used to measure these subscales (n=4) and

by the lower participation rate.

To determine which subscales of the MSLQ

differ significantly between the four clusters a

MANOVA was performed. Since the four clusters

differ in size, a GT2 Hochberg was chosen to

calculate the post-hoc analysis. The results show that

only the scales of self-efficacy and peer learning

differ among the four different clusters.

Cluster 1 students have a high self-efficacy

(M=5.03, SD=1.09) while cluster 4 shows a low

self-efficacy (M=4.35, SD=.72). Cluster 4 students

also show a strong preference for peer learning

(M=3.52, SD=1.28), as do students in cluster 2

(M=3.55, SD=1.34). On the other hand cluster 3

students tend to dislike learning with peers (M=2.49,

SD=1.16). The occurrence of the four different

clusters can, to some extent, be explained by the

dispositional data. To be more specific, the learning

dispositions that show a significant difference

between the four clusters are the tendency to

(dis)like learning with peers and the sense of

competence on the subject matter.

Table 3: Reliability of the subscales of the MSLQ

(n=103).

Reliability

(Cronbach’s Alpha)

Subscale

Motivation

Intrinsic goal orientation .45

Extrinsic goal orientation .51

Task value .76

Self-efficacy .90

Learning strategies

Critical thinking .76

Metacognitive self-regulation .65

Peer learning .72

To establish if the different patterns in the use of

learning resources and subsequently cluster

membership lead to differences in student

performance, an ANOVA was performed with

cluster membership as the factor variable and the

final grade as the dependent variable. A GT2

Hochberg performed the post-hoc analysis since the

clusters differ in size. The results of the ANOVA

can be found in table 4.

Results of the ANOVA showed that students in

cluster 1 and 2 have significant lower course

performance than students in cluster 3 and 4. There

is no significant difference in course performance

between students in cluster 3 and cluster 4.

Proﬁling Student Behavior in a Blended Course - Closing the Gap between Blended Teaching and Blended Learning

Table 4: Average score on the assessment for the different

clusters.

Cluster

number

N 1 2

1 103 4.04

2 143 4.44

3 186 5.16

4 84 5.41

Note: α = 0.05

4 DISCUSSION

The current study explores the different usage

patterns by students of (digital) learning resources. It

tries to establish the causes for these differences by

using dispositional data and determines the effect of

these different usage patterns on student

performance.

Results indicate that there are four usage patterns

of the learning resources as defined by a two-step

cluster analysis. These differences in user patterns

show similarities with previous determined profiles

of technology uses in a blended learning setting: the

no-users (cluster 1), superficial users (cluster 2),

selective active users (cluster 3) and intensive active

users (cluster 4) (Lust et al., 2013; Kovanović et al.,

2015). However, our results revealed that cluster 3

students show a clear preference for online learning

activities and avoid face-to-face activities and are

hence called the selective online users.

When adding dispositional data, gathered by the

MSLQ, to the four clusters we see some emerging

patterns that could explain the causes for differences

in the use of (digital) learning resources. The no-

users in cluster 1 are characterized by a high score

on the subscale of self-efficacy, which may indicate

that they tend to overestimate their performance at

the beginning of the course and they are confident

they will do well. This overestimation presumably

leads them to decide against attending face-to-face

lectures or using the online assessment tools to

determine if they master the subject matter.

The superficial users, cluster 2, are a more

balanced group showing a moderate activity on the

use of all learning resources. Although the MSLQ

showed no significant difference in the subscales for

this group, they have the lowest score of the four

clusters on the subscale extrinsic goal orientation.

Their desire to do well in this course is less evident

compared to the other clusters. This lack of desire

reflects in their superficial use of learning resources:

they use most learning resources in a modest way,

just enough to get by but eventually they fail the

course.

The selective online users in cluster 3 tend to

dislike peer learning. Their tendency to avoid their

peers reflects in their behaviour to compensate their

lecture attendance with online recordings. They

show a slight preference for open essay questions

relative to multiple-choice questions. This usage

pattern reflects a mastery approach towards although

their lack of lecture attendance would suggest

otherwise, as indicated by Wiese and Newton

(2013). They suggest that students with a surface

learning strategy tend to use learning technologies as

a substitute for other learning activities. However,

current research shows that students with a mastery

approach do substitute face-to-face lectures with

online recordings of these lectures.

The intensive active users in cluster 4 visit the

face-to-face lectures most frequently and are

distinguished by a low level of self-efficacy. A low

level of self-efficacy suggests they are insecure

about their performance in the course. They

primarily visit the face-to-face to find reassurance

via the lecturers or their peers. This need for

reassurance is reflected in their use of formative

digital quizzes, in which they prefer to use the

multiple-choice questions above the short essay

questions. They have a need to assess and reflect on

their progress and performance.

In the current research dispositional data play

only a minor role in explaining the differences

between usages of the different learning resources.

This is in contrast with Tempelaar et al. (2015) who

found that learning disposition data serves as a good

proxy at the start of the course for predictive

modelling. The students in the research of

Tempelaar et al., (2015) are more diverse, often with

an international background.

Differences in the use of learning recourses do

have an impact on student performance. Learning

analytics is often used to predict student

performance and to model learning behaviour

(Verbert et al., 2012) but more important is its

purpose to detect undesirable learner behaviour

during a course and adapt the blended course design

so the probability that these behaviours occur is

reduced and redirected. For example, students in

cluster 1 tend to overestimates their skills, resulting

in an underuse of the learning resources. This cluster

would benefit from an educational design that allows

students to gain insight into their own

overestimation.

One of the claimed advantages of blended

learning is that students gain control over their own

learning path and take responsibility for their own

CSEDU 2016 - 8th International Conference on Computer Supported Education

learning (Lust et al., 2013; Orton-Johnson, 2009).

This research shows that students display variety in

the use of learning resources and are designing their

own learning paths or creating their own blends.

However, while these different learning paths do

reflect control of the student, these self-composed

learning paths do not necessary lead to better course

performance. The teacher centred approach, in

which the ‘right’ technology and learning path for

the student have been chosen supplement the course

(Oliver and Trigwell, 2005), is not well embedded in

the current educational design of blended learning,

which implies that using a specific learning

technology is the learner’s decision (Lust et al.,

2011). A more student-centred approach contains an

embedded design of these learning technologies in

which the design either addresses or avoids these

individual differences and consequently redirects

unwanted behaviours. Students need certain

guidance in how to combine learning resources into

an effective learning strategy (Inglis et al., 2011)

since many students don’t seem to master the

metacognitive skills required to control their

learning (Lust et al., 2011) and subsequently do not

choose the learning resources that are the most

effective for them.

The use of data from online learning—learning

analytics—supplemented with dispositional data

gives valuable information about how and why

students use certain learning resources in a blended

course. The use of dispositional data confirms

recommendations made by Shum and Crick (2012)

wherein they conclude that learning analytics

research should be contextualized with a broader set

of indicators.

4.1 Limitations of Current Research

This research uses contextualized data for learner

data analysis in a blended learning setting. However,

even when this context is added, it still reduces the

use of learning resources to visits, clicks and scores

on questionnaires. Research on blended learning

using learning analytics should focus on learning

and ask questions like “What did people learn from

attending this lecture?” rather than, “Did people

attend this lecture?”

Another limitation of the current research is the

known calibration and inaccuracy problems with

self-reports about study tactics (Winne and

Jamieson-Noel, 2002). Students often consider

themselves as self-regulated learners while the

tactics they use to regulate their learning are

ineffective. Moreover, even within a single course

these self-reports about regulation of learning differ

as a function of the task before them (Winne, 2006).

5 CONCLUSIONS

This study showed that there are distinct patterns

between students, which reflects the differences in

use of learning resources in a blended learning

setting. These distinct patterns cause a gap between

blended teaching and learning, with different

patterns leading to differences in student

performance. These distinct patterns can be partially

explained by learning dispositions: student

motivations and their metacognitive ability to

regulate learning. Especially the subscales self-

efficacy and peer learning show significant

differences between different groups of students.

Students with a low self-efficacy have a tendency to

engage in all the learning resources and choose face-

to-face lectures over recorded lectures. Students with

high self-efficacy are confident they will do well in

the course, which causes them to hardly use the

learning resources. Students with a low sense of peer

learning tend to choose lecture recordings over face-

to-face lectures. They use these as a substitute for

lecture attendance.

Although the majority of the subscales of the

MSLQ do not show a significant difference between

the four groups of students, they provide us with

new insights in the gap between blended teaching

and blended learning. The suggestion of Shum and

Crick (2012) to combine learner data with learner

dispositions seems to lead to new insights into why

students do or do not use certain learning

technologies and what the consequences of these

(un)conscious choices are in relation to student

performance.

This research shows that when designing a

blended learning course, the individual differences

in the use of learning resources needs to considered,

but moreover it supports the finding that students

needs specific guidance in the determine what is the

‘right’ (digital) learning resource(s) that supports

their learning.

REFERENCES

Bliuc, A., Goodyear, P. and Ellis, R. (2007). Research

focus and methodological choices in studies into

students' experiences of blended learning in higher

education. The Internet and Higher Education, 10(4),

pp.231-244.

Proﬁling Student Behavior in a Blended Course - Closing the Gap between Blended Teaching and Blended Learning

Bos, N., Groeneveld, C., van Bruggen, J. and Brand-

Gruwel, S. (2015). The use of recorded lectures in

education and the impact on lecture attendance and

exam performance. Br J Educ Technol, p.n/a-n/a.

Chiu, T., Fang, D., Chen, J., Wang, Y. and Jeris, C.

(2001). A robust and scalable clustering algorithm for

mixed type attributes in large database environment.

Proceedings of the seventh ACM SIGKDD

international conference on Knowledge discovery and

data mining - KDD '01.

Ellis, R., Goodyear, P., Calvo, R. and Prosser, M. (2008).

Engineering students' conceptions of and approaches

to learning through discussions in face-to-face and

online contexts. Learning and Instruction, 18(3),

pp.267-282.

George-Walker, L. and Keeffe, M. (2010). Self-

determined blended learning: a case study of blended

learning design. Higher Education Research &

Development, 29(1), pp. 1-13.

Henderson, M., Selwyn, N. and Aston, R. (2015). What

works and why? Student perceptions of ‘useful’ digital

technology in university teaching and learning. Studies

in Higher Education, pp.1-13.

Inglis, M., Palipana, A., Trenholm, S. and Ward, J. (2011).

Individual differences in students' use of optional

learning resources. Journal of Computer Assisted

Learning, 27(6), pp.490-502.

Kovanović, V., Gašević, D., Joksimović, S., Hatala, M.

and Adesope, O. (2015). Analytics of communities of

inquiry: Effects of learning technology use on

cognitive presence in asynchronous online

discussions. The Internet and Higher Education, 27,

pp.74-89.

Long, P. and Siemens, G. (2011). Penetrating the Fog:

Analytics in Learning and Education. Educause

Review, (46).

Lust, G., Elen, J. and Clarebout, G. (2013). Students’ tool-

use within a web enhanced course: Explanatory

mechanisms of students’ tool-use pattern. Computers

in Human Behavior, 29(5), pp.2013-2021.

Lust, G., Vandewaetere, M., Ceulemans, E., Elen, J. and

Clarebout, G. (2011). Tool-use in a blended

undergraduate course: In Search of user profiles.

Computers & Education, 57(3), pp.2135-2144.

Oliver, M. and Trigwell, K. (2005). Can 'Blended

Learning' Be Redeemed?. elea, 2(1), p.17.

Orton-Johnson, K. (2009). ‘I've stuck to the path I'm

afraid’: exploring student non-use of blended learning.

British Journal of Educational Technology, 40(5),

pp.837-847.

Pintrich, P., Smith, D., Garcia, T. and Mckeachie, W.

(1993). Reliability and Predictive Validity of the

Motivated Strategies for Learning Questionnaire

(Mslq). Educational and Psychological Measurement,

53(3), pp.801-813.

Porter, W., Graham, C., Bodily, R. and Sandberg, D.

(2016). A qualitative analysis of institutional drivers

and barriers to blended learning adoption in higher

education. The Internet and Higher Education, 28,

pp.17-27.

Porter, W., Graham, C., Bodily, R. and Sandberg, D.

(2016). A qualitative analysis of institutional drivers

and barriers to blended learning adoption in higher

education.

The Internet and Higher Education, 28,

pp.17-27.

Shum, S. and Crick, R. (2012). Learning dispositions and

transferable competencies. Proceedings of the 2nd

International Conference on Learning Analytics and

Knowledge - LAK '12.

Tempelaar, D., Rienties, B. and Giesbers, B. (2015). In

search for the most informative data for feedback

generation: Learning analytics in a data-rich context.

Computers in Human Behavior, 47, pp.157-167.

Verkroost, M. J., Meijerink, L., Lintsen, H., and Veen, W.

(2008). Finding a balance in dimensions of blended

learning. International Journal on ELearning, 7, 499-

522.

Verbert, K., Drachsler, H., Manouselis, N., Wolpers, M.,

Vuorikari, R. and Duval, E. (2011). Dataset-driven

research for improving recommender systems for

learning. Proceedings of the 1st International

Conference on Learning Analytics and Knowledge -

LAK '11.

Vermunt, J. D. H. M. (1992) Leerstijlen en sturen van

leerprocessen in het hoger onderwijs: naar

procesgerichte instructie in zelfstanding denken

[Learning styles and regulation of learning in higher

education: towards process-oriented instruction in

autonomous thinking] (Amsterdam, Lisse: Swets &

Zeitlinger).

Wiese, C. and Newton, G. (2013). Use of Lecture Capture

in Undergraduate Biological Science Education. cjsotl,

pp.1-24.

Winne, P. (2006). How Software Technologies Can

Improve Research on Learning and Bolster School

Reform. Educational Psychologist, 41(1), pp.5-17.

Winne, P. and Jamieson-Noel, D. (2002). Exploring

students’ calibration of self reports about study tactics

and achievement. Contemporary Educational

Psychology, 27(4), pp.551-572.

Zacharis, N. (2015). A multivariate approach to predicting

student outcomes in web-enabled blended learning

courses. The Internet and Higher Education, 27,

pp.44-53.

CSEDU 2016 - 8th International Conference on Computer Supported Education